Commit Graph

  • cdc550da9a Revert the docker changes to be inline with the original freqtrade image richardjozsa 2022-08-26 09:59:17 +02:00
  • d31926efdf Added Base4Action richardjozsa 2022-08-25 21:40:16 +02:00
  • 53d46a0385 align max_entry_position_adjustment behavior of backtesting to live Matthias 2022-08-25 20:36:17 +02:00
  • 3199eb453b reduce code for base use-case, ensure multiproc inherits custom env, add ability to limit ram use. robcaulk 2022-08-25 19:05:51 +02:00
  • 1fd223c815 Update --prepend help string Matthias 2022-08-25 17:03:41 +02:00
  • 05ccebf9a1 automate eval freq in multiproc robcaulk 2022-08-25 12:29:48 +02:00
  • 94cfc8e63f fix multiproc callback, add continual learning to multiproc, fix totalprofit bug in env, set eval_freq automatically, improve default reward robcaulk 2022-08-25 11:46:18 +02:00
  • f2a356a80c Fix some imports Matthias 2022-08-25 07:08:58 +02:00
  • 6636f17e0f Simplify usage of amount_to_contract precision Matthias 2022-08-25 07:08:22 +02:00
  • 9e48e6a40b Update docs about precision limits in backtesting Matthias 2022-08-25 06:46:57 +02:00
  • 205ab26e92 Remove TODO in test Matthias 2022-08-24 20:54:22 +02:00
  • 70df037690 Improve test precision Matthias 2022-08-24 20:44:48 +02:00
  • 3e786a9b8b added example configs Timothy Pogue 2022-08-24 22:44:22 -06:00
  • d474111a65 Renamed to external signals, controller class refactored Timothy Pogue 2022-08-24 22:42:29 -06:00
  • 592373f096 Remove pairlist waiting, add .db files to .gitignore Timothy Pogue 2022-08-24 18:30:30 -06:00
  • 32faad9333 Fix backtest calculation problem with DCA Matthias 2022-08-24 20:36:08 +02:00
  • d1bee29b1e improve default reward, fix bugs in environment robcaulk 2022-08-24 18:32:40 +02:00
  • a61821e1c6 remove monitor log robcaulk 2022-08-24 16:33:13 +02:00
  • bd870e2331 fix monitor bug, set default values in case user doesnt set params robcaulk 2022-08-24 16:32:14 +02:00
  • c0cee5df07 add continual retraining feature, handly mypy typing reqs, improve docstrings robcaulk 2022-08-24 12:54:02 +02:00
  • b708134c1a switch multiproc thread count to rl_config definition robcaulk 2022-08-23 15:06:57 +02:00
  • b26ed7dea4 fix generic reward, add time duration to reward robcaulk 2022-08-23 14:58:38 +02:00
  • 280a1dc3f8 add live rate, add trade duration robcaulk 2022-08-23 09:44:44 +02:00
  • f9a49744e6 add strategy to the freqai object robcaulk 2022-08-22 19:15:56 +02:00
  • a2a4bc05db Fix the state profit calculation logic richardjozsa 2022-08-22 18:06:33 +02:00
  • 29f0e01c4a expose environment reward parameters to the user config robcaulk 2022-08-21 20:33:09 +02:00
  • d88a0dbf82 add sb3_contrib models to the available agents. include sb3_contrib in requirements. robcaulk 2022-08-21 19:58:36 +02:00
  • 8b3a8234ac fix env bug, allow example strat to short robcaulk 2022-08-21 19:43:39 +02:00
  • 8cd4daad0a Feat/freqai rl dev (#7) mrzdev 2022-08-21 17:43:40 +02:00
  • 3eb897c2f8 reuse callback, allow user to acces all stable_baselines3 agents via config robcaulk 2022-08-20 16:35:29 +02:00
  • 4b9499e321 improve nomenclature and fix short exit bug robcaulk 2022-08-19 11:04:15 +02:00
  • 4baa36bdcf fix persist a single training environment for PPO sonnhfit 2022-08-19 01:49:11 +07:00
  • f95602f6bd persist a single training environment. robcaulk 2022-08-18 16:07:19 +02:00
  • 5d4e5e69fe reinforce training with state info, reinforce prediction with state info, restructure config to accommodate all parameters from any user imported model type. Set 5Act to default env on TDQN. Clean example config. robcaulk 2022-08-18 13:02:47 +02:00
  • 7962a1439b remove keep low profit sonnhfit 2022-08-18 17:53:52 +07:00
  • 81b5aa66e8 make env keep current position when low profit sonnhfit 2022-08-18 17:37:26 +07:00
  • 45218faeb0 fix coding convention sonnhfit 2022-08-18 17:01:04 +07:00
  • d55092ff17 Docker building update, and TDQN repair with the newer release of SB+ richardjozsa 2022-08-17 17:31:27 +02:00
  • 74e4fd0633 ensure config example can work with backtesting RL robcaulk 2022-08-17 12:58:29 +02:00
  • b90da46b1b improve price df handling to enable backtesting robcaulk 2022-08-17 12:51:14 +02:00
  • 2080ff86ed 5ac base fixes in logic MukavaValkku 2022-08-17 08:36:10 +03:00
  • 16cec7dfbd fix save/reload functionality for stablebaselines robcaulk 2022-08-16 12:18:06 +02:00
  • 0475b7cb18 remove unuse code and fix coding conventions sonnhfit 2022-08-16 09:30:35 +07:00
  • d60a166fbf multiproc TDQN with xtra callbacks MukavaValkku 2022-08-15 22:39:33 +03:00
  • dd382dd370 add monitor to eval env so that multiproc can save best_model robcaulk 2022-08-15 18:56:53 +02:00
  • 69d542d3e2 match config and strats to upstream freqai robcaulk 2022-08-15 18:27:48 +02:00
  • e5df39e891 ensuring best_model is placed in ram and saved to disk and loaded from disk robcaulk 2022-08-15 18:08:20 +02:00
  • bf7ceba958 set cpu threads in config robcaulk 2022-08-15 18:01:15 +02:00
  • 57c488a6f1 learning_rate + multicpu changes MukavaValkku 2022-08-15 18:35:41 +03:00
  • 48bb51b458 example config added MukavaValkku 2022-08-15 14:41:24 +03:00
  • b1fc5a06ca example config added MukavaValkku 2022-08-15 14:33:08 +03:00
  • 6d8e838a8f update tensorboard dependency sonnhfit 2022-08-15 22:07:42 +07:00
  • acf3484e88 add multiprocessing variant of ReinforcementLearningPPO robcaulk 2022-08-15 13:46:12 +02:00
  • cf0731095f type fix MukavaValkku 2022-08-15 14:23:00 +03:00
  • 1c81ec6016 3ac and 5ac example strategies MukavaValkku 2022-08-15 14:20:57 +03:00
  • 13cd18dc9a PPO policy change + verbose=1 MukavaValkku 2022-08-15 14:05:01 +03:00
  • 926023935f make base 3ac and base 5ac environments. TDQN defaults to 3AC. robcaulk 2022-08-15 12:13:37 +02:00
  • 096533bcb9 3ac to 5ac MukavaValkku 2022-08-15 12:45:08 +03:00
  • 718c9d0440 action fix MukavaValkku 2022-08-15 12:29:44 +03:00
  • 9c78e6c26f base PPO model only customizes reward for 3AC robcaulk 2022-08-15 11:24:57 +02:00
  • 6048f60f13 get TDQN working with 5 action environment robcaulk 2022-08-15 11:11:16 +02:00
  • d4db5c3281 ensure TDQN class is properly named robcaulk 2022-08-15 10:29:33 +02:00
  • 91683e1dca restructure RL so that user can customize environment robcaulk 2022-08-15 10:26:44 +02:00
  • ecd1f55abc add rl module sonnhfit 2022-08-14 21:26:34 +07:00
  • 70b25461f0 add rl dependency sonnhfit 2022-08-14 20:47:58 +07:00
  • 9b895500b3 initial commit - new dev branch MukavaValkku 2022-08-14 16:24:20 +03:00
  • cd3fe44424 callback function and TDQN model added MukavaValkku 2022-08-13 20:05:21 +03:00
  • 01232e9a1f callback function and TDQN model added MukavaValkku 2022-08-13 18:48:58 +03:00
  • 8eeaab2746 add reward function MukavaValkku 2022-08-12 20:25:13 +03:00
  • ec813434f5 ReinforcementLearningModel MukavaValkku 2022-08-08 19:12:49 +03:00
  • 2f4d73eb06 Revert "ReinforcementLearningModel" MukavaValkku 2022-08-08 19:10:03 +03:00
  • c1e7db3130 ReinforcementLearningModel MukavaValkku 2022-08-08 19:04:53 +03:00
  • 05ed1b544f Working base for reinforcement learning model robcaulk 2022-08-08 15:41:16 +02:00
  • a6d78a8615 initialize Since parameter properly Matthias 2022-08-22 21:18:02 +02:00
  • fe7108ae75 Convert amount to contracts before comparing for close Matthias 2022-08-22 20:48:02 +02:00
  • 78b161e14c add contract_size to database Matthias 2022-08-22 20:32:55 +02:00
  • 6036018f35 Extract contracts_to_amount and amount_to_contracts to standalone functions Matthias 2022-08-22 20:28:33 +02:00
  • 1b0f37a93c Fix documentation typo Matthias 2022-08-22 20:17:04 +02:00
  • 5f38a574ce Add okx broker id Matthias 2022-08-22 20:23:19 +02:00
  • 5ce1c69803 Improve DBSCAN epsilon identification (#7269) th0rntwig 2022-08-22 19:57:20 +02:00
  • 60ba921f56 Merge pull request #7282 from freqtrade/mem-leak-fix Matthias 2022-08-22 19:36:52 +02:00
  • ac42c0153d deprecate indicator_max_period_candles, automatically compute startup candles for FreqAI backtesting. robcaulk 2022-08-22 18:19:07 +02:00
  • 96d8882f1e Plug mem leak, add training timer robcaulk 2022-08-22 13:30:30 +02:00
  • f55d5ffd8c Don't fail when --strategy-path is not a valid directory. Matthias 2022-08-22 09:18:51 +00:00
  • 914b6247e4 Merge pull request #7278 from freqtrade/dependabot/pip/develop/ccxt-1.92.52 Matthias 2022-08-22 08:41:52 +02:00
  • da87e9cbb3 Merge pull request #7275 from freqtrade/dependabot/pip/develop/types-requests-2.28.9 Matthias 2022-08-22 08:41:34 +02:00
  • 484b147a89 Merge pull request #7277 from freqtrade/dependabot/pip/develop/time-machine-2.8.1 Matthias 2022-08-22 07:13:05 +02:00
  • 4fa01548f6 Remove old var from strategy interface Timothy Pogue 2022-08-21 22:49:42 -06:00
  • 6f5478cc02 DataFrame transmission, strategy follower logic Timothy Pogue 2022-08-21 22:45:36 -06:00
  • 015be770c3 ccxt now defaults to base volume for all markets Matthias 2022-08-22 06:42:14 +02:00
  • 93d2f7fc85 types-requests bump pre-commit Matthias 2022-08-22 06:37:26 +02:00
  • 7844157a90 Merge pull request #7276 from freqtrade/dependabot/pip/develop/jsonschema-4.14.0 Matthias 2022-08-22 06:31:29 +02:00
  • 6e046884af Merge pull request #7273 from freqtrade/dependabot/pip/develop/fastapi-0.79.1 Matthias 2022-08-22 06:25:35 +02:00
  • a784f63e9a Merge pull request #7274 from freqtrade/dependabot/pip/develop/mkdocs-material-8.4.1 Matthias 2022-08-22 06:24:22 +02:00
  • ff9ed1abad Bump ccxt from 1.92.20 to 1.92.52 dependabot[bot] 2022-08-22 03:02:25 +00:00
  • 354d3c0cda Bump time-machine from 2.7.1 to 2.8.1 dependabot[bot] 2022-08-22 03:02:10 +00:00
  • eeb177110e Bump jsonschema from 4.9.1 to 4.14.0 dependabot[bot] 2022-08-22 03:02:03 +00:00
  • 70848a258d Bump types-requests from 2.28.8 to 2.28.9 dependabot[bot] 2022-08-22 03:01:59 +00:00
  • 3958e53aaa Bump mkdocs-material from 8.4.0 to 8.4.1 dependabot[bot] 2022-08-22 03:01:55 +00:00
  • dfa7d1fc27 Bump fastapi from 0.79.0 to 0.79.1 dependabot[bot] 2022-08-22 03:01:51 +00:00